Data analysis methods

Data analysis methods include: comparative analysis, grouping analysis, predictive analysis, funnel analysis, AB test analysis, quadrant analysis, formula disassembly, feasible region analysis, and 28/20 analysis method, hypothetical analysis method.

Data analysis methods include: comparative analysis, group analysis, predictive analysis, funnel analysis, AB test analysis, quadrant analysis, formula Dismantling method, feasible region analysis method, 28/20 analysis method, hypothetical analysis method.

1. Comparative analysis method: Contrastive analysis method refers to reflecting the changes in the quantity of things through the comparison of indicators, which is a commonly used method in statistical analysis. Common contrasts include horizontal contrast and vertical contrast.

Horizontal comparison refers to the comparison of different things at a fixed time. For example, the comparison of the prices of goods purchased by users of different levels at the same time, the comparison of the sales volume and profit margin of different goods at the same time.

Vertical comparison refers to the change of the same thing in the time dimension, for example, month-on-month, year-on-year and fixed-base ratio, that is, the comparison of this month’s sales with the previous month’s sales, the January sales of this year and the previous month’s sales Comparison of sales in January of the year, comparison of monthly sales of this year with average sales of the previous year, etc.

Using the comparative analysis method, we can make effective judgments and evaluations on the size, level, speed, etc. of the data.

2. Grouping analysis method: Grouping analysis method refers to dividing the overall data into different parts according to certain indicators according to the nature and characteristics of the data, analyzing its internal structure and interrelationships, so as to understand the nature of things. Development Law. According to the nature of the indicators, the grouping analysis method is divided into attribute indicator grouping and quantitative indicator grouping. The so-called attribute indicators represent the nature and characteristics of things, such as name, gender, education level, etc. These indicators cannot be calculated; while the data represented by data indicators can be calculated, such as a person's age, salary income, etc. The grouping analysis method is generally used in conjunction with the comparative analysis method.

3. Predictive analysis method: Predictive analysis method is mainly based on current data to judge and predict future data change trends. Predictive analysis is generally divided into two types: one is based on time series prediction, for example, predicting sales in the next three months based on past sales performance; the other is regression prediction, which is based on the interaction between indicators. Predictions based on causal relationships, for example, predicting the products a user may purchase based on their web browsing behavior.

4. Funnel analysis method: Funnel analysis method is also called process analysis method. Its main purpose is to focus on the conversion rate of a certain event in important links. It is commonly used in the Internet industry. For example, there are many important links between a user's browsing of card information, submission of application, bank review and card approval, and finally user activation and use, and the number of users in each link is getting smaller and smaller, thus forming a funnel. Using the funnel analysis method, the business side can pay attention to the conversion rate of each link and monitor and manage it. When the conversion rate of a certain link is abnormal, the process can be optimized in a targeted manner and appropriate measures can be taken to improve business indicators.

5. AB test analysis method: AB test analysis method is actually a comparative analysis method, but it focuses on comparing two groups of samples with similar structures, A and B, and analyzes their differences based on the sample index values. . For example, for the same function of an App, different styles and page layouts are designed, and pages of two styles are randomly assigned to users. Finally, the pros and cons of different styles are evaluated based on the user's browsing conversion rate on the page. Understand user preferences to further optimize products.

In addition, in order to do a good job in data analysis, readers also need to master certain mathematical foundations, such as the concepts of basic statistics (mean, variance, mode, median, etc.), dispersion and variability measures (range, quartile, interquartile range, percentile, etc.), data distribution (geometric distribution, binomial distribution, etc.), as well as the basis of probability theory, statistical sampling, confidence intervals and Hypothesis testing and other contents make the data analysis results more professional through the application of relevant indicators and concepts.

6. Quadrant analysis method: The X-axis is the click rate from left to right, and the Y-axis is the conversion rate from bottom to top, forming 4 quadrants. This is the quadrant analysis we want to talk about. Law.

Find the corresponding data labeling points for the click-through rate and conversion rate of each marketing activity, and then classify the effect of this marketing activity into each quadrant. The four quadrants represent different effect evaluations.

7. Formula dismantling method: The so-called formula dismantling method is to use a formula to express the influencing factors of a certain indicator. For example, the influencing factor of daily sales is the sales of each commodity, and find the influencing factors. Finally, the influencing factors need to be dismantled.

8. Feasible region analysis method: Feasible region analysis is actually a self-established data analysis model that continuously corrects and adjusts the scope of the feasible region based on specific data to effectively evaluate business indicators.

9. Twenty-eight analysis method: The eight-eight rule is opposite to the long-tail theory. The twenty-eighth rule tells us that you should pay attention to the top users, that is, the 20% of users or products that can generate 80% of the revenue. The long tail theory tells us to pay attention to the long tail effect, which is the remaining 20% ​​of the income.

10. Hypothesis analysis method: Simple understanding, the hypothesis method is a data analysis method that assumes a quantification among the multiple variables that affect the results based on known result data, and reversely deduce the process.

The above is the detailed content of Data analysis methods. For more information, please follow other related articles on the PHP Chinese website!